Unravelling an optical extreme learning machine

Extreme learning machines (ELMs) are a versatile machine technique that can be seamlessly implemented with optical systems. In short, they described as network of hidden neurons random fixed weights and biases, generate complex behaviour in response to an input. Yet, despite the success physical implementations ELMs, there is still lack fundamental understanding about their implementations. This work makes use media implement ELM introduce ab-initio theoretical framework support experimental implementation. We validate proposed framework, particular, by exploring correlation between rank outputs, H , its generalization capability, thus shedding new light into inner workings ELMs opening paths towards future technological similar principles.